frequency pattern of turbulent ﬂow and sediment …where sediment transport is patchy and...

Hydrol. Earth Syst. Sci., 16, 147–156, 2012www.hydrol-earth-syst-sci.net/16/147/2012/doi:10.5194/hess-16-147-2012© Author(s) 2012. CC Attribution 3.0 License.

Hydrology andEarth System

Sciences

Frequency pattern of turbulent flow and sediment entrainmentover ripples using image processing

A. Keshavarzi1,*, J. Ball2, and H. Nabavi1

1Water Department, Shiraz University, Shiraz, Iran2School of Civil & Environmental Engineering, FEIT, University of Technology Sydney,Broadway, NSW, 2007, Sydney, Australia* now at: School of Civil & Environmental Engineering, FEIT, University of Technology Sydney,Broadway, NSW, 2007, Sydney, Australia

Correspondence to:A. Keshavarzi ([email protected])

Received: 13 July 2011 – Published in Hydrol. Earth Syst. Sci. Discuss.: 17 August 2011Revised: 8 December 2011 – Accepted: 9 December 2011 – Published: 13 January 2012

Abstract. River channel change and bed scourings aresource of major environmental problem for fish and aquatichabitat. The bed form such as ripples and dunes is the resultof an interaction between turbulent flow structure and sed-iment particles at the bed. The structure of turbulent flowover ripples is important to understand initiation of sedimententrainment and its transport. The focus of this study is themeasurement and analysis of the dominant bursting eventsand the flow structure over ripples in the bed of a channel.Two types of ripples with sinusoidal and triangular formswere tested in this study. The velocities of flow over the rip-ples were measured in three dimensions using an AcousticDoppler Velocimeter with a sampling rate of 50 Hz. Thesevelocities were measured at different points within the flowdepth from the bed and at different longitudinal positionsalong the flume. A CCD camera was used to capture 1500 se-quential images from the bed and to monitor sediment move-ment at different positions along the bed. Application of im-age processing technique enabled us to compute the numberof entrained and deposited particles over the ripples. Froma quadrant decomposition of instantaneous velocity fluctu-ations close to the bed, it was found that bursting eventsdownstream of the second ripple, in Quadrants 1 and 3, weredominant whereas upstream of the ripple, Quadrants 2 and4 were dominant. More importantly consideration of theseresults indicates that the normalized occurrence probabili-ties of sweep events along the channel are in phase with thebed forms whereas those of ejection events are out of phasewith the bed form. Therefore entrainment would be expectedto occur upstream and deposition occurs downstream of the

ripple. These expectations were confirmed by measurementof entrained and deposited sediment particles from the bed.These above information can be used in practical applicationfor rivers where restoration is required.

1 Introduction

The entrainment and transport of sediment particles in rivers,natural stream and coastal area is a significant component ofmany environmental degradation problems. Many aquaticecosystems and fish habitats in which needs to be restoredhave problems with channel bed changes and may led to theproblems such as degradation, changes in channel forms andsedimentation. In such rivers ecosystems, restoring the orig-inal morphology and other physical characteristics is nec-essary for the rehabilitation of hydrodynamics of aquaticecosystems. Studies for example by Murphy and Ran-dle (2003) indicated that channel widening is possible onlywith a concerted restoration plan for flow and land manage-ment actions.

Ripples also may create more benefits to aquatic habi-tat resources due to their presence causing a pool habitat tobe created and maintained. Fish communities in the riversand streams are quite sensitive to the availability of stablepools and scour hole volume (Arlinghaus et al., 2002; Arm-strong et al., 2003). Therefore, our understanding of thehydrodynamics of flow structure with links to river restora-tion needs to be considered.

Published by Copernicus Publications on behalf of the European Geosciences Union.

brought to you by COREView metadata, citation and similar papers at core.ac.uk

provided by Crossref

https://core.ac.uk/display/194959374?utm_source=pdf&utm_medium=banner&utm_campaign=pdf-decoration-v1

148 A. Keshavarzi et al.: Frequency pattern of turbulent flow and sediment entrainment over ripples

The hydrodynamics of aquatic life or eco-hydraulics hasbecome recently the interest of many new researches in lastdecade. This new area links the environmental fluid mechan-ics disciplines and ecology together to define flow-organisminteractions at different scales. Nikora (2010) highlighted thehydrodynamics of aquatic ecosystems or ecohydraulics as anemerging tendency in modern science which create a newarea and discipline that needs an integration of several dif-ferent disciplines, offering a useful example of a systematicdevelopment of an integrative discipline in the environmen-tal area. Nikora (2010) concluded that that Hydrodynam-ics of Aquatic Ecosystems will provide a missing researchplatform that will synchronize and enhance flow studies inaquatic ecology and will also provide a solid biophysical ba-sis for ecohydraulics which has been formed as an appliedresearch area based on largely empirical or semi-empiricalapproaches.

In most natural streams, the bed is not flat due to high ve-locity of flow. The high flow velocity produces higher shearstress than shear stress at threshold of initiation of sedimentmotion and makes bed changes in various forms such as rip-ples, dunes and anti-dunes (Knighton, 1998). Ripples anddunes are the result of a combination of high scouring anddeposition processes at the bed of rivers and streams and aremajor causes of channel change.

The formation of bed topography is the result of compli-cated interactions between fluid and sediment particles alongthe bed. Bed forms and the geometry of ripples are a func-tion of bed roughness, median diameter of sediment particlesand flow characteristics such as shear stress, separation andFroude number (Mogridge et al., 1994). Ripple geometryand its interaction with flow structure has been studied bymany investigators (for example, Bagnold, 1946; Carstens etal., 1969; Mogridge et al., 1994; Yalin, 1977; Khelifa andOuellet, 2000; Miller and Komar, 1980; and Nielsen, 1992).In these studies, the geometry of bed ripples was found tobe function turbulent flow structure and shear stress param-eters (Hurthet et al., 2007, 2011; Thorn et al., 2009). Raud-kivi (1997) pointed out that the ripples and vortices withinthe shear layer are affected by the flow depth, velocity distri-bution and shear stress on the bed. Despite more than threedecades of investigation, there is still insufficient informationto characterize ripple-flow interaction in adequate detail andover a range of turbulent flow conditions.

One area where this lack of information exists is in appli-cation of image processing for particle entrainment and anal-ysis of the turbulence characteristics and flow structure overripples. Nevertheless, there has been some research into thisaspect. To study the flow structure over the ripples, Sajjadiet al. (1996) found that the vortices that form in the lee ofripples are important for the entrainment of sediment parti-cles. Keshavarzi and Ball (1999) used image processing tech-nique to record entrained and deposited particles over flat bedand found that there is an intermittent nature for particles en-trainment and deposition over the bed. Venditti et al. (2005)

used high resolution super-VHS video system to monitor thedevelopment of the sand bed over a flat bed and observedthat defect initiation occurs at relatively low flow strengths,where sediment transport is patchy and sporadic. Lajeunesseet al. (2010) used high speed video imaging system to recordthe trajectories of the moving particles over flat bed and tomeasure their velocity and observed that entrained particlesexhibit intermittent motion composed of the succession ofperiods of flight and rest. Bennett and Best (1995, 1996)conducted a series of experiment over fixed bed ripples andcompared the spatial structure of flow over fixed ripples toreveal the contrasts in the dynamics of the flow separationzone over ripples. Kostaschuck and Church (1993), Julienand Klassen (1995), Kostaschuck and Villard (1996), Car-ling et al. (2000), Kostaschuck (2000) and Colombini andStocchino (2011) have conducted field and laboratory stud-ies and concluded that the shear related coherent structureswere very important in bed form development and bed stabil-ity. Consideration of the results from these studies indicatesthat the turbulence characteristics have a direct influence onsediment entrainment and the ripple geometry.

Analysis of the turbulence characteristics is based on theconcept of the bursting phenomenon which was initially in-troduced by Kline et al. (1967) as a means of describing thetransfer of momentum between the turbulent and laminar re-gions near a boundary. Four alternative types of burstingevents have been identified with each of these types havingdifferent effects on the mode and rate of sediment transport(Bridge and Bennett, 1992). Particle entrainment from thebed is closely correlated to the sweep and ejection events(Thorne et al., 1989; Nelson et al., 1995; Drake et al., 1988;Nakagwa and Nezu, 1978; Grass, 1971; Keshavarzi and Ball,1997, 1999). The contribution of sweep and ejection eventshas been found to be more important than outward and in-ward interactions. Additionally, sweep and ejection eventsoccur more frequently than outward and inward interactions(Nakagwa and Nezu, 1978; Thorne et al., 1989; Keshavarziand Ball, 1997) with the average magnitude of the shearstress during a sweep event being much higher than the timeaveraged shear stress (Keshavarzi and Ball, 1997). A num-ber of studies, for example, Offen and Kline (1975) and Pa-panicolaou et al. (2002) investigated the characteristics ofthe bursting process and its effect on particle motion. Fur-thermore, to define the cycle of bursting events, Perona etal. (1998) proposed a simple equation for definition of a typ-ical bursting oscillation. Consideration of these researchesled Yen (2002) to point out the necessity to incorporate thebursting process into the modelling of turbulent flow and sed-iment transport. Jafari Mianaei and Keshavarzi (2008, 2010)found that at the stoss side of ripples, ejection and sweepevents and at the lee side of the ripple outward interactionand inward interaction events were dominant. The studies byOjha and Mazumder (2008) showed that the ratio of shearstress for sweep and ejection events along the dunes variedin an oscillatory pattern at the near bed region, whereas such

Hydrol. Earth Syst. Sci., 16, 147–156, 2012 www.hydrol-earth-syst-sci.net/16/147/2012/

A. Keshavarzi et al.: Frequency pattern of turbulent flow and sediment entrainment over ripples 149

Table 1. Experimental flow conditions.

Test Flow Flow No. of Type of Ripples Ripples Velocity measurement height Froude ReynoldsNo. Rate Depth ripples ripples wave height (mm) Number Number

(l s−1) (mm) length (mm)(mm)

1 14.60 170 2 sinusoidal 250 30 5 0.095 20 8712 18.48 145 2 sinusoidal 250 30 5, 10, 15, 20, 25, 30, 50, 60 0.153 26 4003 18.50 146 2 sinusoidal 250 30 5, 10, 15, 20, 25, 30, 50, 60 0.151 26 4284 20.3 138 2 triangular 200 30 5 0.180 28 9805 22.5 146 2 triangular 250 30 5 0.183 32 1206 25 163 2 triangular 300 30 5 0.173 356 97

Upstream Energy Dissipater Control Gate

Glass Wall V-Notch

Figure 1- A schematic plan of experimental setup

Fig. 1. A schematic plan of experimental setup.

patterns disappear towards the outer flow. Also they foundthat along the dune length, sweep events contribute to shearstress generation. Termini and Sammartano (2009) investi-gated the effects of the variation of bed roughness conditionson the vertical distribution of frequency of the occurrence ofejection and sweep events and concluded that the occurrenceof sweep events increases as the bed roughness increases.

However, in spite of the importance of coherent structuresand its importance to entrainment of sediment particles overthe ripples, their characteristics have not been completely un-derstood. More importantly, the phase probabilities of thesweep and ejection events from the stoss side of the crest andat the lee side of the crest have not been investigated. Pre-sented in this paper, are the results of an investigation intothe phase probability of bursting events and the number ofentrained sediment particles over ripples. An image process-ing technique was used to compute the number of entrainedand deposited sediment particles over the bed and for com-parison with the phase probabilities.

2 Material and methods

The experiments were carried out in a non-recirculating glassflume located in the hydraulics laboratory at Shiraz Univer-sity. The 15.50 m long glass flume has a rectangular cross-section with base width of 0.70 m and height 0.6 m. Thelongitudinal slope of the flume was set at 0.0005. The flowrate was measured using a pre-calibrated 90◦ V-notch weir

λ=250 mm

H=30 mm

Figure 2- A schematic of ripple dimension Fig. 2. A schematic of ripple dimension.

located at the end of the flume and an electromagnetic flowmeter at inlet pipe. An adjustable gate was installed at thedownstream end of the flume for the adjustment of flow depthand velocity. A schematic diagram of the experimental facil-ity is shown in Fig. 1.

In the study reported herein, experimental measurementswere carried out for two different types of ripples includ-ing sinusoidal and triangular shapes. The sinusoidal and tri-angular ripples were made with wavelengths 200, 250 and300 mm and the height of 30 mm. An example photo ofthe ripples dimensions considered in this study is shown inFig. 2. The flume bottom was covered with a layer of bedmaterial, consisting of sand particles with median size (D50)of 0.62 mm.

The experimental tests were performed with fixed bed andmobile bed conditions. For the fixed bed, conditions werevery close to the threshold of sediment entrainment and forthe mobile bed, conditions were at initiation of sedimentmotion and therefore only a low number of sediment parti-cles were in motion. Hence the entrainment process did notchange the ripple form.

The flow conditions of experimental tests performed areshown in Table 1. The flow velocity was measured in threedimensions using an Acoustic Doppler Velocimeter (Micro-ADV) at 128 points within the flow. Velocity measurementswere made at eight points within the flow namely 5, 10,15, 20, 25, 30, 50 and 60 mm from the bed at 16 differentsections over sinusoidal ripples along the flume as shown inFig. 3a. Figure 3b shows bed form configuration of triangular

www.hydrol-earth-syst-sci.net/16/147/2012/ Hydrol. Earth Syst. Sci., 16, 147–156, 2012


a)

A B

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

30 mm 30 mm

25 100 150 200 250 300 350 400 450 500 550 625 X (mm)

b)

11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10

λ=200, 250 & 300 mm

30 mm

Figure 3- Sections and points of velocity measurement a) the sinusoidal ripple; b) the triangular ripple

(a)

a)

A B

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

30 mm 30 mm

25 100 150 200 250 300 350 400 450 500 550 625 X (mm)

b)

11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10

λ=200, 250 & 300 mm

30 mm

Figure 3- Sections and points of velocity measurement a) the sinusoidal ripple; b) the triangular ripple

(b)

Fig. 3. Sections and points of velocity measurement(a) the sinusoidal ripple;(b) the triangular ripple.

ripples and measured positions over the ripples for velocitymeasurement and image capturing in mobile bed condition.More information for specification of triangular ripples ispresented by Jafari Mianaei and Keshavarzi (2008).

The ADV was operated on a pulse-to-pulse coherentDoppler shift to provide three velocity components at a rateof 50 Hz, in which was the maximum sampling frequencyfor instrument with no conditional frequency sampling. Theacoustic sensor consisted of one transmitting transducer andthree receiving transducers. The receiving transducers weremounted on short arms around the transmitting transducerat 120◦ azimuth intervals. The downwards pointing ADVbeams travelling through the water focused at a measuringpoint located 50 mm below the transducer. Therefore, min-imum disturbance to the flow is expected at sampling vol-ume. The output signal passed through a processing modulerequired to evaluate the Doppler shifts. The data acquisitionsoftware provided real-time display of the data in graphicaland tabular forms. Velocities were measured for 120 s at eachpoint, therefore a total 6000 velocity data were collected foreach direction. Data acquisition started after achieving rec-ommended Signal/Noise Ratio (SNR) and correlation coeffi-cient in three dimensions. According to the ADV manufac-turer SonTek (2001) no calibration was necessary while theaccuracy of measurement was within±1.0 %.

To understand 2-D characteristics of turbulent flow priorto investigating the 3-D characteristics of turbulence, thehorizontal and vertical velocity components are analysed inthis study. This approach is consistent with previous stud-ies using quadrant analysis of bursting processes by Kline etal. (1967), Grass (1971), Nakagawa and Nezu (1977, 1978),

Bridge and Bennett (1992), Nelson et al. (1995), Nezu andNakagawa (1993) and Ojha and Mazumder (2008).

3 Results and discussion

3.1 Quadrant decomposition of velocity fluctuations

The bursting process consists of four categories of event;these categories are defined by the quadrant of the event. Asshown in Fig. 4, the events are:

– outward interaction eventu′ > 0, v′ > 0

– ejection eventu′ < 0, v′ > 0,

– inward interaction eventu′ < 0, v′ < 0 and

– sweep eventu′ > 0, v′ < 0.

The velocity fluctuationsv′ andu′ are defined as variationsfrom the time-averaged (mean) velocities in the longitudinaland vertical directions, (u andv), respectively. Algebraically,they are defined by

u′= u − u and v′

= v − v (1)

where;

u =1

n

n∑i=1

ui and v =1

n

n∑i=1

vi . (2)

n being is the number of instantaneous velocity samples.



Zone III Inward Interaction

u′<0, v′<0

Zone IV Sweep

u′>0, v′<0

Zone I Outward Interaction

u′>0, v′>0

Zone II Ejection

u′<0, v′>0

u′

v′>

Figure 4- Four classes of bursting events and their associated quadrants

Fig. 4. Four classes of bursting events and their associated quad-rants.

The time-averaged instantaneous shear stressτ ′ (Reynoldsshear stress) at each point of flow is defined as:

τ ′= −ρ u′ v′ (3)

whereρ is flow density. As shown in Fig. 5, the distributionof the instantaneous velocities is influenced significantly bypresence of ripples.

3.2 Contribution probability of coherent flow andbursting events

Based on two dimensional velocity fluctuations, the occur-rence probability of the bursting events for each quadrant isdefined as;

Pk =nk

N(4)

N =

4∑k=1

nk (5)

wherePk is the occurrence probability of an event in a quad-rant,nk is the number of occurrences of each event,N is thetotal number of events and the subscript represents the indi-vidual quadrants (k = 1 ... 4). Using the above equations, theprobability of each quadrant was computed at each point offlow within the depth. The contributions of coherent struc-tures, such as the sweep (quadrant IV) and ejection (quad-rant II) events, to momentum transfer have been extensivelystudied through quadrant analyses and probability analysesbased on two-dimensional velocity information. Using simi-lar techniques, the contributions of the four events to the en-trainment and motion of sediment particles were determinedfrom the experimental measurements. Shown in Figs. 6 and 7are the quadrant analysis of the frequencies of bursting eventsalong the flume for different depth from the bed.

Section 1 Section 2 Section 3

Section 4


Section 8


Section 12


Section 16

Figure 5- Distribution of instantaneous velocity fluctuations at 20 mm from the bed for 16 sections along the flume

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

u′

v′

Fig. 5. Distribution of instantaneous velocity fluctuations at 20 mmfrom the bed for 16 sections along the flume.

From a quadrant analysis (Figs. 6 and 7) it was found thatat a level of 5 mm from the bed downstream of the second rip-ple crest, quadrants 1 and 3 were dominant when comparedto quadrants 2 and 4. This can be interpreted as an expecta-tion that sedimentation should occur at this location. How-ever, upstream of the ripple crest, quadrants 2 and 4 weremore dominant than quadrants 1 and 3. Therefore entrain-ment would be expected to occur at this location. These ex-pectations are confirmed by measuring sediment particles atthe bed over the ripples along the bed.

The studies by Jafari Mianaei and Keshavarzi (2008, 2010)indicate that at the stoss side of ripples, the time-averagedinstantaneous shear stress of ejection and sweep events weredominant to the outward interaction and inward interactionevents and at the lee side of the ripple it was vice versa. Inother studies for example by Ojha and Mazumder (2008) it isshown that the ratio of time-averaged shear stress for sweepand ejection events along the dunes vary in an oscillatorypattern at the near bed region, whereas such pattern seemsto disappear towards the outer flow. Termini and Sammar-tano (2009) concluded that the occurrence of sweep eventsincreases as the bed roughness increases. The study by Per-ona et al. (1998) was a motivation to definition of burstingcycle in a simple model rather than a complex model. There-fore, in this study to find the relative dominance factor patternof sweep and ejection events along the ripples the ratio of theoccurrence probability of the sweep event is normalized by



a)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

b)

30 mm Ripple

Geometry

Ripple Geometry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

c)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

(a)

a)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

b)

30 mm Ripple

Geometry

Ripple Geometry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

c)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

(b)

a)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

b)

30 mm Ripple

Geometry

Ripple Geometry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

c)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

(c)d)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

e)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

f)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

(d)d)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

e)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

f)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

(e)

d)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

e)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

f)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

(f)

g)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

h)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

Figures 6- Cyclic phase pattern for different classes of bursting events over the ripples along the flume at: a) 5 mm from the bed; b) 10 mm from the bed; c) 15 mm from the bed; d) 20mm from the bed; e) 25 mm from the bed; f) 30mm from the bed; g) 50mm from the bed; h) 60mm from the bed

(g)

g)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

h)

05

101520253035404550

0 10 20 30 40 50 60 70X (cm)

Con

trib

utio

n Pr

obab

ility

(%)

Q1

Q2

Q3

Q4

h

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sections

Figures 6- Cyclic phase pattern for different classes of bursting events over the ripples along the flume at: a) 5 mm from the bed; b) 10 mm from the bed; c) 15 mm from the bed; d) 20mm from the bed; e) 25 mm from the bed; f) 30mm from the bed; g) 50mm from the bed; h) 60mm from the bed

(h)

Fig. 6. Cyclic phase pattern for different classes of bursting events over the ripples along the flume at:(a) 5 mm from the bed;(b) 10 mmfrom the bed;(c) 15 mm from the bed;(d) 20 mm from the bed;(e) 25 mm from the bed;(f) 30 mm from the bed;(g) 50 mm from the bed;(h) 60 mm from the bed.

the occurrence probability of the ejection. Mathematically,this ratio is defined as;

RPSE=P4

P2(6)

whereP4 is the occurrence probability of the sweep event,P2 is the occurrence probability of the ejection events andRPSE is normalized occurrence probability.

Values of the normalised occurrence probability (RPSE)were determined at 5, 10, 15, 20, 25, 30, 50 and 60 mm fromthe bed surface at multiple locations along the flume. Fromthese values of RPSE, a depth average value was determined.

Shown in Fig. 8a and b are the average of RPSE and the bedprofile along the flume. Consideration of these figures indi-cates that the normalized occurrence probabilities of sweepevent are in phase with the bed forms whereas the normalisedoccurrence probabilities of ejection events are out of phasewith the bed profile. Hence, it is concluded that the sweepand ejection events are in a synchronized cyclic pattern withthe bed form.



Table 2. Ratio of entrained sediment particles to the deposited particles (RNED) for triangular ripples.

Section 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Tests

4 1.06 1.03 1.04 1.06 1.08 1.08 1.02 0.99 0.96 1.04 1.06 1.06 1.05 1.08 1.03 1.10 1.045 0.93 1.04 1.03 1.02 1.08 1.09 0.95 0.93 0.90 1.10 1.08 1.16 1.06 1.08 1.06 1.22 1.026 1.06 0.95 1.06 1.04 1.09 1.07 1.04 0.96 0.99 1.02 1.05 0.99 1.04 1.01 0.99 0.94 1.02

a)

b)

c)

d)

(a)

a)

b)

c)

d)

(b)

a)

b)

c)

d)

(c)

a)

b)

c)

d)

(d)

e)

f)

g)

h) Figure 7: Distribution of probability of four quadrants; a: zone 1, b: zone 2, c: zone 3 and d: zone 4 (Test 2); e: zone 1, f: zone 2, g: zone 3 and h: zone 4 (Test 3).

(e)

e)

f)

g)


(f)

e)

f)

g)


(g)

e)

f)

g)


(h)

Fig. 7. Distribution of probability of four quadrants;(a) zone 1,(b) zone 2,(c) zone 3 and(d) zone 4 (Test 2);(e) zone 1,(f) zone 2,(g) zone 3 and(h) zone 4 (Test 3).



a)

0.60.70.80.9

11.11.21.31.41.5

0 2 4 6 8 10 12 14 16 18

Sections

RPS

E

Ripple form

b)

0.60.70.80.91.01.11.21.31.41.5

0 2 4 6 8 10 12 14 16 18Sections

RPS

E

Ripple form

Figure 8: The ratio of phase probability of sweep to ejection events (RPSE) over ripples a) the sinusoidal ripple form; b) the triangular ripple form

(a)

a)

0.60.70.80.9

11.11.21.31.41.5

0 2 4 6 8 10 12 14 16 18

Sections

RPS

E

Ripple form

b)

0.60.70.80.91.01.11.21.31.41.5

0 2 4 6 8 10 12 14 16 18Sections

RPS

E

Ripple form

Figure 8: The ratio of phase probability of sweep to ejection events (RPSE) over ripples a) the sinusoidal ripple form; b) the triangular ripple form

(b)

Fig. 8. The ratio of phase probability of sweep to ejection events (RPSE) over ripples(a) the sinusoidal ripple form;(b) the triangular rippleform.

Figure 9: Two samples of sequential captured images with the difference image between two sequential images

(a)

Figure 9: Two samples of sequential captured images with the difference image between two sequential images

(b)

Fig. 9. Two samples of sequential captured images with the difference image between two sequential images.

3.3 Number of entrained and deposited particles overthe ripples

In this study, to understand sediment entrainment at the bedand over the ripples, an image processing technique was usedto compute number of deposited and entrained sediment par-ticles at the different wavelength and different points overthe ripples. Artificial ripples are used here to make the pos-sibility of particle motion. A CCD camera is used to capturesequential images in 60 s and then the images are simulta-neously recorded in the camera. Therefore, with a rate of25 frames per second a total of 1500 images were recorded.Two samples of sequential recorded image are shown inFig. 9.

To find the number of entrained and deposited particlesan image analysis concept is used to compute sequential im-ages and to produce an image highlighting the differences.In the images differences, the black points represent parti-cles which have been entrained and white points representparticles which are deposited. As a result, from the producedimage, the numbers of entrained and deposited particles werecounted. An example of the difference between two imagesis shown in Fig. 9. The detail of this image processing tech-nique is described by Keshavarzi and Ball (1999). The ra-tio of the number of entrained sediment particles to the de-posited sediment are defined as;

RNED =Number of entrained sediment particles

Number of deposited sediment particles. (7)

0.80.90.91.01.01.11.11.21.2

0 2 4 6 8 10 12 14 16 18

Sections

RN

ED

Figure 10: Ratio of entrained particles to deposited particles (RNED) over ripples

Fig. 10. Ratio of entrained particles to deposited particles (RNED)over ripples.

Table 2 shows the ratio of entrained/deposited number ofparticles over the ripples for different experimental tests.Figure 10 shows the variation of RNED over the ripplesalong the flume. The results indicated that the ratio of en-trained/deposited sediment particles (RNED) is very muchcorrelated with the ratio of the probability of sweep/ejectionevents (RPSE) over the ripples.

An unstated goal of the project reported herein was vali-dation of the experimental methodology for ascertaining therelationship between the turbulent flow structures and ini-tiation of sediment motion when flow over ripples occurs.The influence of the number and the spacing of the ripplesjoin on these relationships remains to be determined. Moreimportantly, the presented frequency pattern in this paper isan initiation of defining a simple model for bursting events.



4 Conclusion

In this study, the flow structure over the ripples was investi-gated experimentally. Consideration of these results showedthat upstream of the first ripple crest, bursting events inquadrants 2 and 4 are dominant, however, downstream ofthe ripple crest, the bursting events in quadrants 1 and 3 aredominant and that sediment deposition occurred downstreamof the ripple crest. Additionally, it was found that theaverage normalized occurrence probabilities of sweep eventare in phase with the bed form whereas the normalizedoccurrence probabilities of ejection event are out of phasewith the bed form. The number of entrained and depositedsediment particles was found also to be in agreement withthe frequency of the bursting events. The above findingis an initiation of defining a simple model for burstingevents, however, it remains to test these findings for the sit-uation of more than two ripples and differing flow conditions.

Acknowledgements.The authors acknowledge Mr Saeed Jafari Mi-anaei, for collecting some parts of laboratory data. Also helpsfrom Reza Keshavarzi for capturing and analysis of the images aregreatly appreciated. Finally the writers would like to thank theanonymous reviewers and Paolo Perona who provided many usefulcomments.

Edited by: P. Perona

References

Arlinghaus, R., Engelhardt, C., Sukhodolov, A., and Wolter, C.:Fish recruitment in a canal with intensive navigation: Implica-tions for ecosystem management, J. Fish Biol., 61, 1386–1402,2002.

Armstrong, J. D., Kemp, P. S., Kennedy, G. J. A., Ladle, M., andMilner, N. J.: Habitat requirements of Atlantic salmon and browntrout in rivers and streams, Fish. Res., 62, 143–170, 2003.

Bagnold, R. A.: Motion of waves in shallow water: interaction be-tween waves and sand bottoms, P. Roy. Soc. Lond. A, 187, 1–18,1946.

Bennett, S. and Best, J.: Mean flow and turbulence structure overfixed, two-dimensional dunes; implications for sediment trans-port and bed form stability, Sedimentology, 42, 491–513, 1995.

Bennett, S. and Best, J.: Mean flow and turbulence structure overfixed ripples and the ripple–dune transition, in: Coherent flow inopen channels, edited by: Ashworth, P., Bennett, S., Best, J., andMcLelland, S., John Wiley, Hoboken, NJ, 281–203, 1996.

Box, G. E. P. and Cox, D. R.: An Analysis of Transformation, J.Roy. Stat. Soc. B, 26, 211–252, 1964.

Bridge, J. S. and Bennett, S. J.: A model for entrainment and trans-port of sediment grains of mixed sizes, shapes and densities, Wa-ter Resour. Res., 28, 337–363, 1992.

Carling, P., Golz, E., Orr, H. G., and Radecki-Pawlik, A.: The mor-phodynamics of fluvial sand dunes in the Ripple-dune transition647 Rhine River, near Mainz, Germany, II. Hydrodynamics andsediment transport, Sedimentology, 47, 253–278, 2000.

Carstens, M. R., Neilson, F. M., and Altinbilek, H. D.: Bed formsgenerated in the laboratory under an oscillatory flow: analyticaland experimental study, US Army Corps of Engineers, CoastalEngineering Research Center, Tech. Memo 28, p.105, 1969.

Colombini, M. and Stocchino, A.: Ripple and dune formation inrivers, J. Fluid Mech., 673, 121–131, 2011.

Drake, T. G., Shreve, R. L., Dietrich, W. E., Whiting, P. J., andLeopold, L. B.: Bed load transport of fine gravel observedby motion-picture photography, J. Fluid Mech., 192, 193–217,1988.

Grass, A. J.: Structural features of turbulent flow over smooth andrough boundaries, J. Fluid Mech., 50, 233–255, 1971.

Hurther, D., Lemmin, U., and Terray, E. A.: Turbulent transport inthe outer region of rough wall open-channel flows: the contribu-tion of Large Coherent Shear Stress Structures (LC3S), J. FluidMech., 574, 465–493, 2007.

Hurther, D., Thorne, P. D., Bricault, M., Lemmin, U., and Barnoud,J. M.: A multi-frequency Acoustic Concentration and VelocityProfiler (ACVP) for boundary layer measurements of fine-scaleflow and sediment transport processes, Coast. Eng., 58, 594–605,2011.

Jafari Mianaei, S. and Keshavarzi, A.: Spatio-temporal variation oftransition probability of bursting events over the ripples at the bedof open channel, J. Stoch. Environ. Res. Risk A., 22, 257–264,2008.

Jafari Mianaei, S. and Keshavarzi, A.: Study of near bed stochas-tic turbulence and sediment entrainment over the ripples at thebed of open channel using image processing technique, J. Stoch.Environ. Res. Risk A., 24, 591–598, 2010.

Julien, P. and Klassen, G.: Sand-dune geometry of large rivers dur-ing floods, J. Hydraul. Eng.-ASCE, 121, 657–663, 1995.

Keshavarzi, A. R. and Ball, J. E.: An analysis of the characteristicsof rough bed turbulent shear stress in an open channel flow, J.Stoch. Hydrol. Hydraul., 11, 193–210, 1997.

Keshavarzi, A. R. and Ball, J. E.: An application of image pro-cessing in the study of sediment motion, J Hydraul. Res., 37,559–576, 1999.

Khelifa, A. and Ouellet, Y.: Prediction of sand ripple geometryunder waves and currents, Journal Waterway Port Coast. OceanEng., 126, 14–22, 2000.

Kline, S. J., Reynolds, W. C., Schraub, F. A., and Runstadler, P. W.:The structure of turbulent boundary layers, J Fluid Mech., 30,741–773, 1967.

Knighton, D.: Fluvial Forms and Processes, Arnold, London, Eng-land, 1998.

Kostaschuck, R.: A field study of turbulence and sediment dynam-ics over subaqueous dunes with flow separation, Sedimentology,47, 519–531, 2000.

Kostaschuck, R. and Church, M.: Macroturbulence generated bydunes, Sediment. Geol., 85, 25–37, 1993.

Kostaschuck, R. and Villard, P.: Flow and sediment transport overlarge subaqueous dunes, Fraser River, Canada, Sedimentology,43, 849–863, 1996.

Lajeunesse, E., Malverti, L., and Charru, F.: Bed load transport inturbulent flow at the grain scale: Experiments and modeling, J.Geophys. Res., 115, F04001,doi:10.1029/2009JF001628, 2010.

Mogridge, G. R., Davies, M. H., and Willis, D. H.: Geometry pre-diction for wave generated bedforms, Coast. Eng., 22, 255–286,1994.


http://dx.doi.org/10.1029/2009JF001628


Miller, M. C. and Komar, P. D.: Oscillation sand ripples generatedby laboratory apparatus, J. Sediment. Petrol., 50, 173–182, 1980.

Murphy, P. J. and Randle, T. J.: Platte River Sediment Transport andRiparian Vegetation Model, Technical Report, US Department ofthe Interior, Bureau of Reclamation, Denver, Co, 2003.

Nakagwa, H. and Nezu, I.: Prediction of the contributions to theReynolds stress from bursting events in open-channel flows, J.Fluid Mech., 80, 99–128, 1977.

Nakagwa, H. and Nezu, I.: Bursting phenomenon near the wall inopen channel flow and its simple mathematical model, Fac EngKyoto University, Japan XL, 40, 213–240, 1978.

Nelson, J. M., Shreve, R. L., McLean, S. R., and Drake, T. G.: Roleof near bed turbulence structure in bed load transport and bedform mechanics, Water Resour. Res., 31, 2071–2086, 1995.

Nezu, I. and Nakagawa, H.: Turbulence in open channel flows,IAHR Monograph, Balkema, Rotterdam, 1993.

Nielsen, P.: Coastal bottom boundary layers and sediment transport,World Scientific, 324 pp., 1992.

Nikora, V.: Hydrodynamics of aquatic systems: An interface be-tween ecology, biomechanics and environmental fluid mechan-ics, River Res. Appl., 26, 367–384, 2010.

Ojha, S. P. and Mazumder, B. S.: Turbulence characteristics of flowregion over a series of 2-D dune shaped structures, Adv. WaterResour., 31, 561–576, 2008.

Offen, G. R. and Kline, S. J.: A proposed model of the burstingprocess in turbulent boundary layers, J. Fluid Mech., 70, 209–228, 1975.

Papanicolaou, A. N., Diplas, P., Evaggelopoulos, N., and Fotopou-los, S.: Stochastic incipient motion criterion for spheres undervarious packing conditions, J. Hydraul. Eng.-ASCE, 128, 369–380, 2002.

Perona, P., Porporato, A., and Ridolfi, L.: A simple experimen-tal equation for the bursting cycle, Phys. Fluids, 10, 3023–3026,1998.

Raudkivi, A.: Ripples of stream bed, J. Hydraul. Eng.-ASCE, 123,58–64, 1997.

Sajjadi, S. G., Aldridge, J. N., and Nicholas, D. J.: Sediment trans-port mechanism over rippled sand beds, in: Estuarine and coastalmodelling, edited by: Spaulding, M. L. and Cheng, R. T., ASCE,New York, 669–680, 1996.

Termini, D. and Sammartano, V.: Sediment flow interactions and ef-fects on bed forms mechanism, 7th ISE & 8th HIC, Chile, 2009.

Thorne, P. D., Williams, J. J., and Heathershaw, A. D.: In situ acous-tic measurements of marine gravel threshold and transport, Sed-imentology, 36, 61–74, 1989.

Thorne, P. D., Davies, J. S., and Bell, P. S.: Observationsand analysis of sediment diffusivity profiles over sandy rip-pled beds under waves, J. Geophys. Res., 114, C02023,doi:10.1029/2008JC004944, 2009.

Venditti, J. G., Church, M. A., and Bennett, S. J.: Bed form ini-tiation from a flat sand bed, J. Geophys. Res., 110, F01009,doi:10.1029/2004JF000149, 2005.

Yalin, M. S.: On the determination of ripple length, Proc. Am. Soc.Civil Eng., 103, 439–442, 1977.

Yen, B. C.: Stochastic inference to sediment and fluvial hydraulics,J. Hydraul. Eng.-ASCE, 128, 365–367, 2002.


http://dx.doi.org/10.1029/2008JC004944

http://dx.doi.org/10.1029/2004JF000149

frequency pattern of turbulent ﬂow and sediment …where sediment transport is patchy and...

Documents